⚡ Productivity
#meetings#notes#tasks#productivity

Meeting Notes to Action Items

Paste your raw meeting notes or transcript into this prompt and get back a clean list of action items — each with an owner, a deadline, and the context it came from. No more decisions evaporating the moment the call ends.

Turns rambling meeting notes into crisp owners, deadlines, and decisions — nothing slips through.

You are a precision meeting analyst. I will paste raw, messy meeting notes. Extract: 1) DECISIONS MADE — each as one clear sentence, 2) ACTION ITEMS — each with an owner, a specific deliverable, and a deadline (infer reasonable ones and mark them as [inferred]), 3) OPEN QUESTIONS — anything unresolved, 4) A 3-line executive summary. Format as clean markdown. If an action item has no clear owner, assign it to the most likely person and flag it. Here are the notes:
More productivity prompts
Try it now ChatGPT ↗ Grok ↗
Example output

See what this prompt creates

AI-generated example output for the 'Meeting Notes to Action Items' prompt
Example output generated by AI with this prompt. Your results will vary by tool and the details you fill in.

What this prompt does

Meetings produce words; work needs tasks. This prompt bridges the gap: feed it messy notes, a transcript, or even bullet fragments, and it extracts every commitment, decision, and follow-up — then structures them with owners and deadlines. What took 30 minutes of “so what did we decide?” becomes a shareable action list in seconds.

How to get the best result

Include who was in the meeting — the prompt assigns owners from the attendee list, which is the whole point. Raw transcripts work best (don't pre-clean them; the model is good at finding signal in rambling). If deadlines were vague (“next week”), it flags them for you to confirm rather than guessing silently.

Common mistakes to avoid

Don't strip out the small talk before pasting — decisions often hide in casual asides (“oh, I'll just handle the invoice then”). And always sanity-check owner assignments: the model infers from context, and a misassigned task is worse than an unassigned one because nobody follows up on it.

Try these variations

Use it on sales calls to extract objections and follow-ups, or on 1-on-1s to track commitments over time — paste four weeks of 1-on-1 notes and ask for patterns. Combine with the inbox zero triage prompt for a full “communications to tasks” system, in the productivity prompts collection.

How to use this prompt

  1. Copy it — hit the copy button above to grab the raw prompt text.
  2. Fill the brackets — replace anything in [BRACKETS] with your own details.
  3. Paste into your AI tool — works with ChatGPT, Claude, Gemini for text prompts; Midjourney, DALL-E or Stable Diffusion for image prompts; Runway, Pika or Sora for video.
  4. Iterate — tweak one bracket at a time until the output is exactly right.

Frequently asked questions

Does it work with auto-generated transcripts from Zoom or Teams?

Perfectly — that's its favorite input. Paste the raw transcript including timestamps; it ignores filler and extracts decisions, tasks, and owners. Messy, unedited transcripts actually work better than over-cleaned notes.

What if no clear owner was mentioned for a task?

The prompt flags ownerless tasks explicitly instead of inventing an assignee. That's deliberate: an unassigned task in a highlighted “needs owner” list gets claimed fast, while a wrongly assigned one gets ignored.

Can it track action items across multiple meetings?

Yes. Paste several meetings' notes together and ask for a consolidated tracker — it merges duplicates, shows what's carried over, and surfaces items that keep getting postponed. Great for weekly project reviews.

How accurate is the extraction?

Very good for explicit commitments (“I'll send the report Friday”), decent for implied ones. Treat the output as a draft: a 2-minute review pass catches the model's occasional over-eager task invention.

Related prompts

Keep exploring the vault.

Explore more categories

More prompt collections worth your time.